5-azacytidine (also known as azacitidine and 4-amino-1-β-D-ribofuranosyl-S-triazin-2(1H)-one; Nation Service Center designation NSC-102816; CAS Registry Number 320-67-2) has undergone NCI-sponsored trials for the treatment of myelodysplastic syndromes (MDS). See Kornblith et al., J. Clin. Oncol. 20(10): 2441-2452 (2002) and Silverman et al., J. Clin. Oncol. 20(10): 2429-2440 (2002). 5-azacytidine may be defined as having a formula of C8H12N4O5, a molecular weight of 244.20 and a structure of

The s-triazine ring of 5-azacytidine has a particular sensitivity to water (see J. A. Beisler, J. Med. Chem., 21, 204 (1978)); this characteristic has made the synthesis of 5-azacytidine a challenge, especially in manufacturing at commercial scale. A number of prior art methods have been developed in order to avoid the use of water; however, these methods all have additional problems that render them undesirable for the production of large-scale batches of 5-azacytidine. For example, Piskala and Sorm teach the following synthesis scheme in (see U.S. Pat. No. 3,350,388; A. Piskala and F. Sorm, Collect. Czech. Chem. Commun., 29, 2060 (1964); and A. Piskala and F. Sorm, Ger. 1922702 (1969), each of which is incorporated herein by reference
in its entirety):
The overall yield of this scheme is 43.3%. This method involves a reactive starting material (isocyanate) with a controlled stereochemistry (1-β configuration). Such a compound cannot be regarded as a starting material. The drawbacks of this scheme include the presence of steps that are difficult to scale-up, the use of benzene as solvent in one step, and the requirement for a deprotection step performed in a closing pressure vessel using dry ammonia. Furthermore, the final 5-azacytidine product was isolated from the reaction mixture by filtration with no further purification; this is not acceptable for the synthesis of an Active Pharmaceutical Ingredient (API) for human use. The addition of further purification steps will further reduce the overall yield.
Winkley and Robins teach an 5-azacytidine synthesis process that relies on the coupling of a “bromosugar” with a silyl derivative of 5-azacytosine (see M. W. Winkley and R. K. Robins, J. Org. Chem., 35, 491 (1970), incorporated by reference in its entirety):

In this procedure, 5-azacytosine was treated with excess hexamethyldisilazane (HMDS) in the presence of catalytic amounts of ammonium sulfate at reflux until a complete solution was generated (TMS=(CH3)3Si). See E. Wittenburg, Z. Chem., 4, 303 (1964) for the general procedure. The excess HMDS was removed by vacuum distillation and the residue was used directly (without further purification) in the coupling with, 2,3,5-tri-O-acetyl-D-ribofuranosyl bromide in acetonitrile. The coupled product was deprotected with methanolic ammonia solution.
There are many significant weaknesses in this procedure. First, the fact that the bromosugar was a mixture of anomers, which means that the final coupled product was also a mixture of anomers. Second, the work-up in the coupling step involved a great many steps, specifically: concentration of the reaction mixture to dryness; treatment of the residue with sodium bicarbonate, water and methanol; removal of the water by co-evaporation with absolute ethanol; extraction of the residue with chloroform twice; and finally the concentration to dryness of the combined chloroform in extract. Third, ammonia in MeOH was used in the deprotection step, which requires the use of a pressure vessel. Fourth, the crude 5-azacytidine was isolated in only a 35% yield. This crude material was then dissolved in warm water and the solution was decolorized with charcoal. Evaporation then gave crystals of 5-azacytidine with a yield 11%. This material was further recrystallized from aqueous ethanol (charcoal). The low recovery during purification can be correlated with the poor anomeric ratio and with the known low stability of 5-azacytidine in water.
Piskala and Sorm also teach the following process for coupling involving the use of a “chlorosugar” (A. Piskala and F. Sorm, Nuci. Acid Chem. 1, 435 (1978), incorporated herein by reference in its entirety):

2,3,5-Tri-O-Benzoyl-D-ribofuranosyl chloride was prepared by saturating a solution of 1-O-acetyl-2,3,5-tri-O-benzoyl-β-D-ribose in CICH2CH2Cl—AcCl with gaseous HCl (with ice-cooling) and then keeping the mixture overnight at room temperature. This procedure is difficult to scale-up with plant equipment due to the special handling requirements of gaseous HCI. Also, the typical α/β ratio in the chlorosugar is unknown, as is the impact of the α/β ratio on the yield and final purity of 5-azacytidine.
Piskala, Fiedler and Sorm teach a procedure for the ribosylation of silver salts of 5-azapyrimidine nucleobases in A. Piskala, P. Fiedler and F. Sorm, Nucleic Acid Res., Spec. Publ. 1, 17 (1975), incorporated herein by reference in its entirety. Specifically, they teach that the ribosylation of the silver salt of 5-azacytosine with 2,4,5-tri-O-benzoyl-D-ribosyl chloride gives 5-azacytidine. This is clearly not a procedure that is amenable to scale up for the large-scale production of 5-azacytidine.
Niedballa and Vorbrüggen teach the procedure that has been used historically for the large-scale synthesis of 5-azacytidine for the above-mentioned NCI-sponsored trials for the treatment of myelodysplastic syndromes. See H. Vorbrüggen and U. Niedballa, Ger. 2,012,888 (1971) and U. Niedballa and H. Vorbrüggen, J. Org. Chem., 39, 3672 (1974), each of which is incorporated herein by reference in its entirety. The procedure involves the following steps:
There are at least three major drawbacks to this procedure. First, and most importantly,
after purification, variable amounts of tin from one batch to another were found in the API. The lack of control of the tin level means that the procedure is not suitable for producing an API for human use. Second, emulsions developed during the workup of the coupling mixture. Indeed, H. Vorbrüggen and C. Ruh-Pohlenz in Organic Reactions, Vol. 55, 2000 (L. A. Paquette Ed., John Wiley & Sons, New York), p 100, have previously noted that silylated heterocycles and protected 1-O-acyl or 1-O-alkyl sugars in the presence of Friedel-Crafts catalysts like SnCI4 often form emulsions and colloids during work-up. The phase separation of the emulsion is slow, so the water-sensitive protected 5-azacytidine was exposed to water for variable periods of time leading to variable amounts of decomposition. Third, a filtration step was performed in order to isolate the insoluble tin salt. Typically, this filtration is very slow, and is likely the reason that variations in the final yield were noted. These problems mean that the process is not conveniently amenable to scale-up.
Vorbrüggen et al. in Chemische Berichte, 114: 1234-1255 (1981) teach the use of certain Lewis acids as Friedel-Crafts catalysts for the coupling of silylated bases with 1-O-acyl sugars. In particular, they teach the coupling of silylated bases with 1-O-acyl sugars in the presence of trimethylsilyl trifluoromethanesulfonate (TMS-Triflate) in 1,2-dichloroethane or acetonitrile. The reaction mixture was then diluted with dichloromethane and the organic phase extracted with ice-cold saturated NaHCO3. The use of this procedure to synthesize 5-azacytidine is not taught or suggested.
Vorbrüggen and Bennua in Chemische Berichte, 114: 1279-1286 (1981) also teach a simplified version of this nucleoside synthesis method in which base silylation, generation of the Lewis acid Friedel-Crafts catalyst, and coupling of the silylated base to the 1-O-acyl sugar takes place in a one step/one pot procedure employing a polar solvent such as acetonitrile. Following reaction, dichloromethane is added, and the mixture is extracted with aqueous NaHCO3. The use of this procedure to synthesize 5-azacytidine is not taught or suggested. Moreover, this one step/one pot reaction is not suitable for the synthesis of 5-azacytidine because the extraction is done is the presence of acetonitrile. Acetonitrile is a polar solvent, and is therefore miscible with water. As a consequence, the protected 5-azacytidine in the acetonitrile is exposed during extraction to the aqueous phase for variable amounts of time, which in turn leads to variable amounts of decomposition of the protected 5-azacytidine.
Thus, there is an unmet need in the field for the provision of a simple, controlled procedure for the synthesis of 5-azacytidine that provides an API that is suitable for use in humans, minimizes the exposure of 5-azacytidine to water, and is amenable to scaling-up for the production of large quantities of 5-azacytidine.